Pages: p. 4
by Mary Shaw and Paul Clements, pp. 31–39. This retrospective on nearly two decades of research examines the software architecture field's maturation by tracing the evolution of its research questions and results. Early qualitative results set the stage for later precision, formality, and automation. Results over the ensuing decades have matured and moved into practice.
by Christian F.J. Lange, Michel R.V. Chaudron, and Johan Muskens, pp. 40–46. To determine how UML is being used in current software architecting and design, the authors surveyed practitioners and analyzed case studies of industry projects. Their results show that UML is used rather loosely and that UML models are often incomplete. This leads to miscommunication and other implementation and maintenance problems. The authors conclude with recommendations and techniques for controlling UML model quality.
by Robert L. Nord and James E. Tomayko, pp. 47–53. Including architecture-centric design and analysis methods in the Extreme Programming framework can help software developers address quality attributes in an explicit, methodical, engineering-principled way. Properly managed, architecture-centric methods can be a cost-effective addition to the software development process and will increase system and product quality.
by Michael Stal, pp. 54–61. Using software patterns and blueprints to express a service-oriented architecture's fundamental principles supports the efficient use of SOA technologies for application development. Software patterns can express almost all architecture principles that span the space of SOA technologies. This architecture-centric approach helps developers understand service-oriented infrastructures and build SOA applications that meet operational and developmental requirements.
by Jacqueline Floch, Svein Hallsteinsen, Erlend Stav, Frank Eliassen, Ketil Lund, and Eli Gjørven, pp. 62–70. Recently, the introduction of software platforms supporting component plug-in and dynamic binding has facilitated adaptation of software systems at runtime. Preserving the properties described by the architecture model during adaptation is an important task. The authors propose a self-adaptation approach that exploits architecture models to reason about and control adaptation at runtime. They can derive runtime models from design models facilitating the developers' task. They developed the approach in the context of mobile computing.
by Alek Radjenovic and Richard Paige, pp. 71–79. Safety-critical systems, also known as high-integrity real-time systems, require architecture description languages that model the entire system rather than specific parts or aspects of it. ADLs pose unique challenges in their construction. Architecture Information Modeling is an ADL developed in collaboration with the safety industry. AIM provides language flexibility, abstraction layering, custom view creation, a design-by-view process, and safety and change control. A case study demonstrates the application of the AIM concepts in a top-down software design.
by Rogério de Lemos, Paulo Asterio de Castro Guerra, and Cecília Mary Fischer Rubira, pp. 80–87. Developers typically address dependability concerns in the late phases of system development. However, two trends are compelling developers to consider dependability earlier, at the architectural level. First, emerging applications are increasingly complex. Second, to address this first trend, developers are increasingly attempting to build dependable systems from existing undependable components. A new architectural approach employs exception handling to represent and analyze fault-tolerant software systems. It partitions architectural elements into normal and exceptional parts, thus promoting a clear separation of concerns regarding how to detect and handle errors.
by Silvia T. Acuña, Natalia Juristo, and Ana M. Moreno, pp. 94–101. The human dimension is a critical factor in organizations, especially in software companies where the production process is essentially intellectual. One key aspect of the software engineering workforce is identifying the people best suited for development roles. Despite its importance, there is little support for this task, leading to unease and misunderstandings among the people involved. A capability-based procedure can aid managers at small to medium-sized software organizations.
by Anthony M. Orme, Haining Yao, and Letha H. Etzkorn, pp. 102–108. Measuring system coupling is a commonly accepted software engineering practice associated with producing high-quality software products. In many application domains, we can assess coupling in ontology-based systems before system development by measuring coupling in ontology data. A proposed set of metrics measures coupling of ontology data in ontology-based systems represented in the Web Ontology Language, a derivative of XML. A real-world study demonstrates the metrics' use in integrating ontologies.